Traditionally, estimation of regional flow using externally detected tracer kinetics has been performed by means of transit time analysis. However, under in vivo conditions, indicator recirculation usually interferes with transit time analysis. This limitation can be overcome by analyzing the residue curves using a model that realistically represents the vascular architecture, and modeling the vascular input function. This approach is sensitive to the temporal resolution of the residue data, the dispersion of the input function and the noise content of the data. The goal of this project is to develop general approaches for estimating regional flow using X-ray CT, MRI, ultrasonic and PET tracers and to define the limitations of these approaches.